()

As the United Kingdom plans its exit from the European Union, dedicating time to fine tune legal and compliance discovery processes with multilingual capabilities will be increasingly important to companies and their outside counsel.

It’s expected that the U.K.’s historic referendum to leave the EU will result in more cross-border litigation, investigations and regulatory oversight for U.S. and U.K.-based businesses with international operations due to additional complexities brought on by the farewell. Consequently, a potential influx of multilingual-document reviews await legal counsel. Here are a few reasons why:

• Brexit could have some impact on the enforceability of contracts governed under English law, and many companies are considering reviews of material contracts where they are likely to include provisions that may be triggered by Brexit.

• For financial institutions, there may be increased regulatory implications necessitating reviews.

• From an intellectual property perspective, there is uncertainty surrounding the Unified Patent Court system, as Brexit signals an end to the U.K.’s involvement in this “one stop shop” system, possibly weakening the attraction of U.K. courts as the venue to settle patent disputes.

• While English is the first choice for EU institutions, no country other than the U.K. has it registered as its primary language. Regulations (by unanimous EU adoption) would have to thus be changed in EU states to maintain the language, challenging the current ubiquity of English within EU government institutions and corporations.

Multi-language reviews are notoriously complicated. So until Brexit becomes more of a reality and can guide data privacy laws, understanding the impact foreign language will have on e-discovery, where the major pitfalls can occur—and how to avoid them—can make legal matters more compliant and cost-effective.

Identify Multilingual Documents Before Starting

Legal teams collecting documents from offices or subsidiaries based in the U.K. should anticipate that their data collections will contain multilingual language documents due to some of the factors discussed above. Otherwise, they’ll invest countless hours and dollars in categorizing documents by search terms, only to later come across documents written in Arabic or Chinese that will require them to start the process anew to account for the multi-language documents.

Many e-discovery review platforms have the capability of detecting foreign languages. The software must be able to handle languages that do not use the Roman alphabet during the processing, search, and analytics phases of the review.

Organizations that use discovery software with predictive analytics that parses language at the sentence level, rather than at the document level, can identify which languages a data set contains as well as the percentage of each language in the collection. Armed with this information, they can choose the right software, tailor workflows, and marshal the necessary human resources for a successful review project.

Understand Complexities of Foreign Languages

Foreign languages alone are complex. Instant messaging, SMS, social media, and other evolving channels—while revolutionizing how we communicate—have simultaneously complicated them more so.

For example, “Y do we O $? LMK UR thoughts,” meaning “Why do we owe money? Let me know your thoughts,” is difficult enough to handle in English. Now you must add the complexity of a foreign language to slang and text speak. Efforts to use search terms will almost certainly not capture the universe of potentially relevant information.

When foreign language combines with chat-speak or other obscure slang, how can a reviewer or computer algorithm detect legally significant innuendo?

Machine translation tools can handle mass volumes of data quickly and cheaply and give counsel a general understanding of the basic premise and relevancy of a document. However, the results will not always be precise, especially when documents contain unfamiliar jargon, nonstandard abbreviations, convoluted legal syntax, or sophisticated financial data that require accurate translation for defensible review. Typically, organizations will need to recruit resources to review these documents in the specific foreign language.

Engage Industry and Domain Experts

Once documents that are likely to be responsive are isolated, the next step is often a foreign-language review for tone, context, or idioms and jargon that machine translation tools can’t assess.

Finding qualified native document reviewers abroad can be challenging. For this reason, many organizations look to service providers that have scalable foreign language and subject-matter experts readily available. The key is to find an organization with a roster of well-trained and managed local reviewers with the requisite skill set.

Language fluency is the bare minimum requirement, and proficiency should be verified through testing. Experience in the nuances of American discovery, privilege concepts, and the industry and subject matter of the underlying legal case or investigation are all highly desirable. More knowledgeable professionals can lend their insight in the development and refinement of search terms and syntax. They can also educate the rest of the review team on regional dialects, geographic-specific business terms, and conversational quirks, both formal and informal.

Service providers must be thoroughly vetted up front to confirm that they apply rigorous recruiting standards, train reviewers thoroughly and supervise them closely. The presence of strong on-site project management teams can also ensure greater consistency and more reliable performance.

Control Review Costs

With the size of today’s data sets, translating and reviewing every document is wasteful and unrealistic. And linear review in any foreign language progresses more slowly than English-language reviews. Additionally, the fees for foreign language review can quickly escalate depending on a set of factors, such as the special skills and qualifications required, the language requiring review, the complexity of the task (i.e., simpler translation or more involved review), and the location of the review.

In any review, the best way to lower the discovery budget is to limit the amount of review time. This occurs in two steps: reducing the number of foreign language documents that require review and then increasing the speed of review.

As discussed earlier, when used at the inception of a project, machine translation can sample foreign language documents so review teams can make more informed decisions about relevance. In some cases, organizations may be tempted to proceed to native translation to gain a more precise understanding before review. However, given the exorbitant cost of human translation, it may make good financial sense for reviews with a high percentage of multi-language documents to have foreign language attorneys first review the documents in their native language and then only translate the documents that are relevant to the case.

To expedite review, organizations should choose a discovery suite with robust analytics tools capable of comparing concepts and relationships across documents regardless of language, so reviewers can target the most critical custodians and identify anomalous behavioral patterns. Keyword filtering is still viable with these documents, but it is critical that discovery specialists collaborate with linguistics experts, who can ensure that search terms account for local cultures, dialects and colloquialisms.

Predictive coding can work with any language, so long as the seed set is coded by a skilled attorney with the requisite knowledge, and can prioritize the documents most likely to be responsive for early review and cull nonresponsive documents. Tools such as email threading, deduplication, and near-duplicate identification can avoid the rework that occurs when multiple copies of identical or similar documents proliferate throughout a data set. Other analytics tools that cluster related documents can provide the context required to hasten understanding and review.

Keep in mind that the rules may change if a document collection contains digitized hard-copy documents. Analytics and machine translation tools are ideal for native electronic files, since they can access both the text and metadata of a document, but they are less effective with scanned files that must undergo optical character recognition.

Adhere to Data Privacy Rules

Until the U.K. formally withdraws from the EU, the U.K.’s Data Protection Act 1998, which implements the EU Data Protection Directive, is still the law of the land. While refining processes to better handle multi-language documents, organizations must still consider a host of issues to remain compliant, including how to set up proper access controls to the secured data; where to conduct the document review (either in-country, on site or elsewhere as allowed); and how to protect sensitive data to comply with privacy rules, such as by redaction. Records that require heightened protection include personally identifying information, payment card information, protected health information and geolocation data.

Many organizations form a cross-functional team that includes their data privacy officer, outside counsel and third-party discovery vendors. The team can then be tasked with the development of a defensible strategy that addresses the issues of legally acceptable data transfer and processing. In addition, organizations concerned about the sanctity of their data and reluctant to transfer their data offsite can opt for a mobile “backpack” solution, which will allow all processing to occur on site, within the organization’s firewall.

Take a Proactive Approach to Compliance

A large number of organizations still take a reactive approach to reviews, reviewing documents (including those in foreign languages) on a case-by-case basis after a legal matter or compliance review has commenced. When documents are parsed out to outside counsel for review, it can complicate a review if different coding designations are made and sensitive or private data is inadvertently disclosed through production.

So, many companies are now adopting analytics systems for ongoing compliance monitoring. Large amounts of data can be filtered to the important facts to help flag potentially “risky” activity on a real-time basis, identify isolated incidents and create alerts to systematic issues for further review (and remediation) by legal and compliance teams—before litigation hits.

Conclusion

Legal and compliance reviews are daunting enough on their own. Add the complications of an event like Brexit, and even the most reliable protocols and tactics for reviewing significant numbers of foreign language documents can snare the savviest counsel. By understanding the obstacles, applying the right technology and recruiting skilled personnel, organizations and their legal counsel can minimize the risks.

As the United Kingdom plans its exit from the European Union, dedicating time to fine tune legal and compliance discovery processes with multilingual capabilities will be increasingly important to companies and their outside counsel.

It’s expected that the U.K.’s historic referendum to leave the EU will result in more cross-border litigation, investigations and regulatory oversight for U.S. and U.K.-based businesses with international operations due to additional complexities brought on by the farewell. Consequently, a potential influx of multilingual-document reviews await legal counsel. Here are a few reasons why:

• Brexit could have some impact on the enforceability of contracts governed under English law, and many companies are considering reviews of material contracts where they are likely to include provisions that may be triggered by Brexit.

• For financial institutions, there may be increased regulatory implications necessitating reviews.

• From an intellectual property perspective, there is uncertainty surrounding the Unified Patent Court system, as Brexit signals an end to the U.K.’s involvement in this “one stop shop” system, possibly weakening the attraction of U.K. courts as the venue to settle patent disputes.

• While English is the first choice for EU institutions, no country other than the U.K. has it registered as its primary language. Regulations (by unanimous EU adoption) would have to thus be changed in EU states to maintain the language, challenging the current ubiquity of English within EU government institutions and corporations.

Multi-language reviews are notoriously complicated. So until Brexit becomes more of a reality and can guide data privacy laws, understanding the impact foreign language will have on e-discovery, where the major pitfalls can occur—and how to avoid them—can make legal matters more compliant and cost-effective.

Identify Multilingual Documents Before Starting

Legal teams collecting documents from offices or subsidiaries based in the U.K. should anticipate that their data collections will contain multilingual language documents due to some of the factors discussed above. Otherwise, they’ll invest countless hours and dollars in categorizing documents by search terms, only to later come across documents written in Arabic or Chinese that will require them to start the process anew to account for the multi-language documents.

Many e-discovery review platforms have the capability of detecting foreign languages. The software must be able to handle languages that do not use the Roman alphabet during the processing, search, and analytics phases of the review.

Organizations that use discovery software with predictive analytics that parses language at the sentence level, rather than at the document level, can identify which languages a data set contains as well as the percentage of each language in the collection. Armed with this information, they can choose the right software, tailor workflows, and marshal the necessary human resources for a successful review project.

Understand Complexities of Foreign Languages

Foreign languages alone are complex. Instant messaging, SMS, social media, and other evolving channels—while revolutionizing how we communicate—have simultaneously complicated them more so.

For example, “Y do we O $? LMK UR thoughts,” meaning “Why do we owe money? Let me know your thoughts,” is difficult enough to handle in English. Now you must add the complexity of a foreign language to slang and text speak. Efforts to use search terms will almost certainly not capture the universe of potentially relevant information.

When foreign language combines with chat-speak or other obscure slang, how can a reviewer or computer algorithm detect legally significant innuendo?

Machine translation tools can handle mass volumes of data quickly and cheaply and give counsel a general understanding of the basic premise and relevancy of a document. However, the results will not always be precise, especially when documents contain unfamiliar jargon, nonstandard abbreviations, convoluted legal syntax, or sophisticated financial data that require accurate translation for defensible review. Typically, organizations will need to recruit resources to review these documents in the specific foreign language.

Engage Industry and Domain Experts

Once documents that are likely to be responsive are isolated, the next step is often a foreign-language review for tone, context, or idioms and jargon that machine translation tools can’t assess.

Finding qualified native document reviewers abroad can be challenging. For this reason, many organizations look to service providers that have scalable foreign language and subject-matter experts readily available. The key is to find an organization with a roster of well-trained and managed local reviewers with the requisite skill set.

Language fluency is the bare minimum requirement, and proficiency should be verified through testing. Experience in the nuances of American discovery, privilege concepts, and the industry and subject matter of the underlying legal case or investigation are all highly desirable. More knowledgeable professionals can lend their insight in the development and refinement of search terms and syntax. They can also educate the rest of the review team on regional dialects, geographic-specific business terms, and conversational quirks, both formal and informal.

Service providers must be thoroughly vetted up front to confirm that they apply rigorous recruiting standards, train reviewers thoroughly and supervise them closely. The presence of strong on-site project management teams can also ensure greater consistency and more reliable performance.

Control Review Costs

With the size of today’s data sets, translating and reviewing every document is wasteful and unrealistic. And linear review in any foreign language progresses more slowly than English-language reviews. Additionally, the fees for foreign language review can quickly escalate depending on a set of factors, such as the special skills and qualifications required, the language requiring review, the complexity of the task (i.e., simpler translation or more involved review), and the location of the review.

In any review, the best way to lower the discovery budget is to limit the amount of review time. This occurs in two steps: reducing the number of foreign language documents that require review and then increasing the speed of review.

As discussed earlier, when used at the inception of a project, machine translation can sample foreign language documents so review teams can make more informed decisions about relevance. In some cases, organizations may be tempted to proceed to native translation to gain a more precise understanding before review. However, given the exorbitant cost of human translation, it may make good financial sense for reviews with a high percentage of multi-language documents to have foreign language attorneys first review the documents in their native language and then only translate the documents that are relevant to the case.

To expedite review, organizations should choose a discovery suite with robust analytics tools capable of comparing concepts and relationships across documents regardless of language, so reviewers can target the most critical custodians and identify anomalous behavioral patterns. Keyword filtering is still viable with these documents, but it is critical that discovery specialists collaborate with linguistics experts, who can ensure that search terms account for local cultures, dialects and colloquialisms.

Predictive coding can work with any language, so long as the seed set is coded by a skilled attorney with the requisite knowledge, and can prioritize the documents most likely to be responsive for early review and cull nonresponsive documents. Tools such as email threading, deduplication, and near-duplicate identification can avoid the rework that occurs when multiple copies of identical or similar documents proliferate throughout a data set. Other analytics tools that cluster related documents can provide the context required to hasten understanding and review.

Keep in mind that the rules may change if a document collection contains digitized hard-copy documents. Analytics and machine translation tools are ideal for native electronic files, since they can access both the text and metadata of a document, but they are less effective with scanned files that must undergo optical character recognition.

Adhere to Data Privacy Rules

Until the U.K. formally withdraws from the EU, the U.K.’s Data Protection Act 1998, which implements the EU Data Protection Directive, is still the law of the land. While refining processes to better handle multi-language documents, organizations must still consider a host of issues to remain compliant, including how to set up proper access controls to the secured data; where to conduct the document review (either in-country, on site or elsewhere as allowed); and how to protect sensitive data to comply with privacy rules, such as by redaction. Records that require heightened protection include personally identifying information, payment card information, protected health information and geolocation data.

Many organizations form a cross-functional team that includes their data privacy officer, outside counsel and third-party discovery vendors. The team can then be tasked with the development of a defensible strategy that addresses the issues of legally acceptable data transfer and processing. In addition, organizations concerned about the sanctity of their data and reluctant to transfer their data offsite can opt for a mobile “backpack” solution, which will allow all processing to occur on site, within the organization’s firewall.

Take a Proactive Approach to Compliance

A large number of organizations still take a reactive approach to reviews, reviewing documents (including those in foreign languages) on a case-by-case basis after a legal matter or compliance review has commenced. When documents are parsed out to outside counsel for review, it can complicate a review if different coding designations are made and sensitive or private data is inadvertently disclosed through production.

So, many companies are now adopting analytics systems for ongoing compliance monitoring. Large amounts of data can be filtered to the important facts to help flag potentially “risky” activity on a real-time basis, identify isolated incidents and create alerts to systematic issues for further review (and remediation) by legal and compliance teams—before litigation hits.

Conclusion

Legal and compliance reviews are daunting enough on their own. Add the complications of an event like Brexit, and even the most reliable protocols and tactics for reviewing significant numbers of foreign language documents can snare the savviest counsel. By understanding the obstacles, applying the right technology and recruiting skilled personnel, organizations and their legal counsel can minimize the risks.